Electronics Guide

The SI Electrical Units

The modern SI defines electrical units through their relationships to fundamental constants:

• Ampere: Defined by fixing the elementary charge e at exactly 1.602176634 × 10⁻¹⁹ coulombs, so that one ampere, one coulomb per second, is a flow of 1/(1.602176634 × 10⁻¹⁹), about 6.241509 × 10¹⁸, elementary charges per second
• Volt: Derived from the ampere, watt, and joule; practically realized using the Josephson effect, which relates voltage to frequency through Planck's constant and the elementary charge
• Ohm: Derived from the volt and ampere; practically realized using the quantum Hall effect, which produces resistance determined by Planck's constant and the elementary charge

These definitions ensure that electrical units are based on constants of nature rather than the properties of any particular material or artifact, providing stability and reproducibility that artifact standards could never achieve.

Conventional Values and Practical Realization

Before 2019, the international metrology community used conventional values of the Josephson constant, K_J-90 = 483,597.9 GHz/V, and the von Klitzing constant, R_K-90 = 25,812.807 Ω, to maintain worldwide consistency in electrical measurements. These conventional values, recommended by the International Committee for Weights and Measures (CIPM) in 1988 and in use from 1 January 1990, differed slightly from the best available measurements of the true constants.

The 2019 SI revision replaced these conventional values with exact values derived from the fixed fundamental constants. Both are exact, although their decimal expansions do not terminate:

• Josephson constant K_J: K_J = 2e/h = 483,597.8484... GHz/V
• von Klitzing constant R_K: R_K = h/e² = 25,812.80745... Ω

This transition introduced small discontinuities in the maintained electrical units. Because K_J is smaller than K_J-90 by about 1.07 parts in 10⁷, the numerical value of a voltage measured with a Josephson standard rose by that amount, and because R_K exceeds R_K-90 by about 1.78 parts in 10⁸, the numerical value of a resistance measured with a quantum Hall standard rose by that amount. In exchange, the practical electrical units became part of the SI, and no further adjustments are needed as measurement techniques improve.

Physics of the Josephson Effect

A Josephson junction consists of two superconductors separated by a thin barrier, typically an insulating oxide layer or a weak link. In the superconducting state, electron pairs (Cooper pairs) can tunnel through this barrier, creating a supercurrent.

When the junction is biased with a DC voltage V, the supercurrent oscillates at frequency f = K_JV. Conversely, when microwave radiation at frequency f is applied, the junction develops DC voltage steps at integer multiples of f/K_J. These steps arise from phase-locking between the internal Josephson oscillation and the external microwave field.

The remarkable feature of this effect is that the voltage depends only on frequency and fundamental constants, with no dependence on junction materials, dimensions, temperature (as long as superconductivity is maintained), or any other extrinsic factor. Two Josephson junctions made from different materials in different laboratories produce identical voltages when irradiated at the same frequency.

Josephson Junction Types

Several junction technologies have been developed for voltage standards:

• SIS Junctions: Superconductor-insulator-superconductor junctions, typically niobium with aluminum oxide barriers, are underdamped and hysteretic; they are the standard type for conventional standards
• SNS Junctions: Superconductor-normal metal-superconductor junctions are overdamped, with a single-valued current-voltage characteristic, which makes them the usual choice for programmable standards
• SINIS Junctions: Superconductor-insulator-normal metal-insulator-superconductor stacks are also non-hysteretic and have been used in programmable arrays

Junction fabrication requires sophisticated thin-film deposition and lithography techniques to achieve the uniformity needed for large arrays.

Conventional Josephson Standards

The voltage of a single step is tiny. At a typical drive frequency of 70 GHz, the first step of one junction is

V = f/K_J = 70 GHz / 483,597.8484 GHz/V ≈ 144.7 µV

Early Josephson standards, built from one or a few junctions, therefore delivered only millivolts, too little for practical calibrations. Useful output voltages require thousands of junctions connected in series.

Conventional 10 V standards contain about 20,000 junctions in series. The junctions rest simultaneously on stable quantized voltage steps, and the total output is the sum of all junction voltages. The microwave radiation is distributed to all junctions through a carefully designed waveguide structure. Because the hysteretic SIS junctions have steps that extend across zero bias current, the whole array can be biased by a single current near zero, but each junction may settle on any of several steps. The operator or control software therefore adjusts the bias until the array settles on a combination of steps that gives the desired total, then fine-tunes the frequency to reach the exact target voltage.

Key characteristics of conventional standards include:

• Output voltage: Typically 1 V or 10 V, determined by junction count and microwave frequency
• Step selection: Adjusted by the operator or by software to reach the desired quantized level; the output cannot be switched quickly from one voltage to another
• Stability: Voltage stable to parts in 10¹¹ when properly operated
• Output capability: Essentially no output current; the array cannot drive a load and is used in null comparisons

Conventional standards serve as primary standards at national metrology institutes and as transfer standards for calibrating secondary standards.

Programmable Josephson Voltage Standards

Programmable Josephson voltage standards (PJVS) represent a significant advancement, enabling rapid output voltage changes and the synthesis of arbitrary waveforms with quantum accuracy.

A PJVS uses overdamped junctions whose steps do not overlap, so a bias current in the right range selects a step unambiguously. The array is divided into independently biased segments, often binary-weighted in size, and each segment can be biased to the +1, 0, or −1 voltage step. By selectively activating different combinations of segments, the total output voltage can be programmed to any value within the array's range in discrete quantum steps.

The number of junctions a given output needs follows from the step size, and so from the drive frequency. A 10 V array needs about 10 V / 144.7 µV ≈ 69,000 junctions at 70 GHz, but about 10 V / 41.4 µV ≈ 242,000 junctions at 20 GHz. Practical PJVS arrays therefore range from several thousand to a few hundred thousand junctions, providing:

• Programmable DC output: Any voltage from zero to the maximum, in increments as small as one junction's step, f/K_J, which is about 41.4 µV at 20 GHz; finer values are reached by adjusting the drive frequency
• Step-approximated waveforms: Sine waves, triangles, and arbitrary shapes synthesized from quantized steps
• Fast settling: Output can change in microseconds, enabling AC calibrations
• Intrinsic accuracy: Every output level has quantum accuracy

PJVS enables calibration of AC voltmeters, RMS-to-DC converters, sampling systems, and other AC instruments against quantum standards, which conventional standards cannot do. The transitions between steps are not themselves quantized, however, so stepwise-approximated waveforms are most accurate at relatively low frequencies, and sampling techniques are often used to exclude the transitions from the measurement. Pulse-driven arrays, known as Josephson arbitrary waveform synthesizers (JAWS), extend quantum-accurate AC synthesis to higher frequencies at smaller amplitudes.

Operating a Josephson Standard

Josephson standards require careful attention to operating conditions:

• Cryogenic cooling: Junctions must be cooled below their superconducting critical temperature, typically 4.2 K using liquid helium or mechanical cryocoolers
• Microwave excitation: Precise frequency control is essential; the microwave source is locked to a laboratory frequency standard or a GPS-disciplined oscillator
• Bias current: DC bias positions the junction on the desired voltage step; must be carefully controlled
• Magnetic shielding: External magnetic fields can affect junction behavior; shielding is essential
• EMI protection: Electromagnetic interference can cause step transitions; careful filtering and shielding are required

Despite these requirements, Josephson standards have become routine tools at national metrology institutes and major calibration laboratories worldwide.

Uncertainty in Josephson Standards

The voltage from an ideal Josephson junction is exactly V = nf/K_J, with zero uncertainty from the junction itself. Practical uncertainty arises from:

• Frequency measurement: Uncertainty in the microwave frequency reference contributes directly; GPS-disciplined oscillators achieve parts in 10¹²
• Step selection: Uncertainty about which step the junction occupies; proper biasing eliminates this concern
• Leakage and offsets: Small currents and thermal EMFs in the measurement circuit add uncertainty
• Comparison method: The null detector and comparison technique introduce additional uncertainty

State-of-the-art Josephson voltage standards achieve combined uncertainties below 1 part in 10¹⁰ for DC voltage comparisons.

Physics of the Quantum Hall Effect

In a two-dimensional conductor at low temperature and high magnetic field, electrons occupy discrete Landau levels. When the Fermi energy lies between Landau levels, the longitudinal resistance vanishes and the Hall resistance plateaus at quantized values.

The quantum Hall plateaus are extraordinarily flat and reproducible. The Hall resistance depends only on fundamental constants, independent of sample material, geometry, temperature (within range), magnetic field (within plateau), and other factors. Different samples made from different materials in different laboratories produce identical resistance values on the same plateau.

The most commonly used plateau is i = 2, giving

R_H = R_K/2 ≈ 12,906.4 Ω

This value, while not a round number, can be compared to standard resistors using precision resistance bridges.

Quantum Hall Devices

Quantum Hall devices are fabricated from materials that support a two-dimensional electron gas (2DEG):

• GaAs/AlGaAs heterostructures: The traditional choice, providing a high-mobility 2DEG at cryogenic temperatures; requires cooling to about 1.5 K or below and magnetic fields above about 5 T, commonly near 10 T
• Silicon MOSFETs: The devices in which von Klitzing discovered the effect; now rarely used for standards, but important in demonstrating the universality of the effect
• Graphene: Epitaxial graphene on silicon carbide, whose widely spaced Landau levels preserve accurate quantization at higher temperatures and lower magnetic fields; increasingly adopted by national metrology institutes

GaAs devices are grown by molecular beam epitaxy or similar techniques, and graphene devices by high-temperature growth on silicon carbide substrates. In both cases lithographic patterning then defines the Hall bar geometry and contacts.

Operating Conditions

Quantum Hall measurements require extreme conditions:

• Temperature: About 1.5 K or below for GaAs devices, typically about 1.3 K in a pumped helium-4 cryostat or about 0.3 K in a helium-3 refrigerator
• Magnetic field: From about 5 T to 14 T depending on the device's carrier density, requiring superconducting magnets
• Current: Typically microamperes to avoid heating and non-equilibrium effects
• Sample quality: High mobility is essential; careful material growth and handling required

Graphene relaxes these requirements considerably. Graphene devices have reproduced R_K/2 to within about 2 parts in 10⁹ at 4.2 K in a field of about 4.5 T, and NIST has operated a graphene standard in a tabletop cryocooler with a 5 T magnet that needs no liquid helium. The quantum Hall effect has also been observed in graphene at room temperature, but only in fields of about 45 T and without metrological accuracy, so room-temperature standards remain a research goal rather than a prospect.

Resistance Measurement and Scaling

The quantum Hall resistance of about 12.9 kΩ must be compared to practical resistance values using precision bridges and scaling techniques:

• Cryogenic current comparators: Use superconducting circuits to compare currents with extreme precision, enabling resistance ratio measurements at parts in 10⁹
• DC comparator bridges: Room-temperature bridges comparing unknown resistors to the quantum Hall standard through intermediate standards
• Resistance scaling: Build-up and build-down techniques using 10:1 or 100:1 ratio measurements to reach practical decade values

The scaling chain from quantum Hall resistance to standard decade values (1, 10, 100, 1,000, and 10,000 ohms) introduces additional uncertainty at each step, though modern techniques keep this contribution small.

Uncertainty in Quantum Hall Standards

Like Josephson standards, quantum Hall devices produce intrinsically accurate values limited mainly by practical measurement considerations:

• Plateau identification: Must verify operation on the intended plateau, not a higher or lower one
• Contact resistance: Non-ideal contacts can introduce errors; four-terminal measurements eliminate lead resistance
• Heating effects: Excessive current can warm the sample and degrade quantization
• Comparison method: Bridge and comparator uncertainties add to the overall budget

The quantization itself is reproducible to about 1 part in 10¹⁰: comparisons of devices made from different materials, such as graphene and GaAs, agree at that level. Calibrating a standard resistor against a quantum Hall standard adds the bridge and the resistor's own instability, and at the best laboratories reaches uncertainties of parts in 10⁹.

Construction and Chemistry

The saturated Weston cell consists of:

• Positive electrode: Mercury covered with mercurous sulfate paste
• Negative electrode: Cadmium amalgam (cadmium dissolved in mercury)
• Electrolyte: Saturated cadmium sulfate solution with excess crystals to maintain saturation
• Container: H-shaped glass vessel with sealed construction

The cell reaction produces an EMF of approximately 1.01864 volts at 20 degrees Celsius. This specific voltage has no particular significance; it simply results from the electrochemistry of the chosen materials.

Temperature Dependence

The EMF of a saturated Weston cell varies with temperature according to a well-characterized polynomial:

E(t) − E(20) = −40.6(t − 20) − 0.95(t − 20)² + 0.01(t − 20)³

where t is the cell temperature in degrees Celsius and E(t) − E(20), the departure from the EMF at 20 degrees Celsius, is in microvolts. The slope is −40.6 µV per degree Celsius at 20 degrees Celsius, about 40 ppm of the cell's EMF per degree, and steepens to about −49 µV per degree Celsius at 25 degrees Celsius. This substantial temperature coefficient requires that standard cells be maintained at controlled temperature, typically in oil baths stabilized to millidegrees: a change of 1 millidegree shifts the EMF by about 0.04 µV, or 4 parts in 10⁸.

Unsaturated Weston cells, which use a lower concentration electrolyte, have much smaller temperature coefficients but are less stable long-term.

Handling and Use

Weston cells are delicate instruments requiring careful handling:

• No current draw: Cells must not supply any significant current; even microamperes can polarize the electrodes and change the EMF
• Temperature stability: Must be maintained at constant temperature during measurements
• Orientation: Should not be tilted or inverted, which can disturb the internal structure
• Transportation: Requires special procedures to avoid shock and temperature excursions
• Connections: Must use high-quality connections to avoid thermoelectric EMFs

When properly maintained, groups of standard cells can maintain their assigned voltage values to within a few parts per million over decades.

Standard Cell Groups

National metrology institutes maintained groups of standard cells to improve stability:

• Mean of group: The average EMF of multiple cells is more stable than any individual cell
• Intercomparison: Regular comparison between cells in the group detects individual drift
• Transport cells: Selected stable cells used for international comparisons
• Working cells: Cells used for routine calibrations, traceable to the primary group

Even with groups, standard cells exhibited slow drift requiring periodic international comparisons to maintain worldwide consistency. The Josephson standard eliminated this drift problem entirely.

Zener Diode References

The avalanche breakdown in a reverse-biased silicon junction provides a voltage reference that has become the workhorse of practical electronics. Temperature-compensated Zener references place an avalanche diode in series with a forward-biased junction, combining the positive temperature coefficient of avalanche breakdown with the negative coefficient of the forward-biased junction to achieve low overall drift. The best devices use a buried, or subsurface, Zener junction, isolated from the surface contamination that causes noise and drift, and some, such as the LM399 and LTZ1000, place a heater and temperature sensor on the same die so that the junction runs at a constant elevated temperature.

Characteristics of precision Zener references include:

• Output voltage: Typically about 6.2 to 7.5 V, where the avalanche and forward-junction temperature coefficients can be balanced
• Temperature coefficient: Below 1 ppm per degree Celsius for premium devices; the heated LM399 is specified at a typical 0.3 ppm per degree Celsius
• Long-term stability: 1 to 10 ppm per year after initial aging
• Noise: Zener breakdown generates significant noise; filtering required for precision applications
• Operating current: Must be controlled; usually includes current regulation circuitry

Bandgap Voltage References

Bandgap references exploit the temperature characteristics of silicon to produce a stable output voltage near 1.2 to 1.25 V. That output sits close to the voltage of about 1.2 V to which a silicon junction's forward voltage extrapolates at absolute zero, a figure related to, but distinct from, the bandgap energy of silicon, which is about 1.12 eV at room temperature.

The bandgap principle combines:

• V_BE of a transistor: A temperature coefficient of approximately −2 mV per degree Celsius
• ΔV_BE: The difference in V_BE between transistors at different current densities, ΔV_BE = (kT/q) ln(J₁/J₂), which is proportional to absolute temperature; for a current-density ratio of 8 it is about 54 mV at 300 K and rises by about 0.18 mV per kelvin
• Weighted sum: Combining these in the right proportion cancels the first-order temperature variation

Modern bandgap references add curvature correction to compensate for higher-order temperature terms, achieving temperature coefficients below 5 ppm per degree Celsius over the full operating range. Bandgap references are small, need only a low supply voltage, and fit in any integrated circuit process, but they are noisier and drift more over time than the best buried-Zener references, so the Zener remains the choice for laboratory standards.

Precision Reference ICs

Integrated circuit references combine reference elements with regulation, trimming, and output amplification:

• Temperature compensation: On-chip temperature sensors and correction circuits minimize drift
• Voltage scaling: Precision resistor networks scale the basic reference to standard output voltages (2.5 V, 5 V, 10 V)
• Laser trimming: Factory adjustment for initial accuracy and temperature coefficient
• Low noise: Careful design and filtering minimize output noise
• Stability: Premium references achieve long-term stability of 1 to 2 ppm per year

The best solid-state references achieve initial accuracy of 0.01% and temperature coefficients below 1 ppm per degree Celsius. Initial accuracy matters little in a laboratory standard, since calibration replaces the nominal value with a measured one; what qualifies a reference as a working standard, once calibrated against a Josephson standard or a traceable transfer standard, is its stability, noise, and temperature coefficient.

Ovenized References

For the highest stability, precision references are operated in temperature-controlled ovens that maintain the reference at constant temperature above ambient:

• Oven temperature: Set above the highest expected ambient temperature, so that the heater never loses control; a lower set point slows aging but narrows the usable ambient range
• Temperature control: Proportional or proportional-integral control maintains temperature within millidegrees
• Warm-up time: A heated reference IC stabilizes within seconds to minutes, but a complete oven may need hours to reach full thermal equilibrium
• Power consumption: Continuous heating required, which is a burden in battery-powered equipment

Ovenized references can achieve effective temperature coefficients as low as about 0.05 ppm per degree Celsius. At that level the limit is set by the control loop and by the temperature coefficients of the resistors that set the oven temperature and bias the reference, rather than by the reference element itself.

Reference Calibration and Traceability

Solid-state references must be calibrated against higher-level standards to establish traceability:

• Initial calibration: Comparison against a Josephson standard or traceable transfer standard establishes the absolute value
• Drift characterization: Multiple calibrations over time establish the drift rate
• Uncertainty assignment: The calibration uncertainty combines comparison uncertainty with expected drift between calibrations
• Recalibration interval: Based on drift rate and accuracy requirements; typically annual for working references

A well-characterized reference with known drift rate can provide accurate voltage values between calibrations by applying a time-based correction.

Wire-Wound Standard Resistors

Traditional standard resistors use precision resistance wire wound on insulating forms:

• Manganin wire: Copper-manganese-nickel alloy, typically about 86 percent copper, 12 percent manganese, and 2 percent nickel, with very low temperature coefficient (near zero at room temperature) and low thermoelectric EMF against copper
• Evanohm: Higher resistivity alloy suitable for high-value standards
• Four-terminal construction: Separate current and voltage terminals eliminate lead resistance from measurements
• Strain-free mounting: Mechanical stress affects resistance; careful mounting minimizes this effect
• Hermetic sealing: Oil-filled or hermetically sealed enclosures protect against humidity and atmospheric pressure effects

Premium wire-wound standards achieve temperature coefficients below 1 ppm per degree Celsius and stability of a few parts per million per year.

Rosa Standard Resistors

The Rosa design, introduced by Edward B. Rosa at the National Bureau of Standards (NBS, now NIST) in 1909, seals a manganin coil in a can filled with mineral oil:

• Hermetic sealing: The earlier Reichsanstalt-type standards were open to the air, and their resistance changed with atmospheric humidity; sealing the coil removed that effect
• Oil immersion: Ensures thermal uniformity and protects the wire from the environment
• Temperature measurement: The temperature of the oil can be measured so that readings can be corrected for the coil's temperature coefficient
• Low self-heating: Large wire gauge and oil cooling minimize resistance change due to measurement current

Thomas-Type Standard Resistors

James L. Thomas of NBS developed his one-ohm standard around 1930 while investigating the small changes in resistance of Rosa-type standards. Thomas-type standards represent the highest stability achieved with wire-wound construction:

• Double-walled enclosure: The coil is hermetically sealed inside a double-walled container, in thermal contact with the inner wall, so that heat from the measurement current escapes readily to the surrounding oil bath
• Annealed manganin: Heavy manganin wire, annealed at high temperature in a vacuum, minimizes internal stress and time-dependent drift
• Stability: Best examples drift less than 0.1 ppm per year
• Temperature coefficient: Adjusted to near zero at the operating temperature

A select group of Thomas-type resistors maintained the United States legal ohm from 1939 until the quantum Hall effect took over that role in 1990. They remain valuable as stable transfer standards.

Film and Foil Resistors

Modern precision resistors use deposited films or bulk foil technology:

• Thin-film resistors: Metal alloy films deposited on ceramic substrates, laser-trimmed for precision; common in integrated precision references
• Bulk metal foil: Precision alloy foil bonded to ceramic; combines excellent stability with low temperature coefficient
• Temperature coefficient: Best foil resistors achieve below 0.5 ppm per degree Celsius
• Load coefficient: Resistance change with applied power; foil resistors excel due to efficient heat spreading

AC-DC Transfer Standards

Resistors used for AC measurements must maintain their value across frequency:

• Skin effect: At high frequencies, current flows near the wire surface, increasing effective resistance
• Inductance: Wire-wound resistors have significant inductance; bifilar winding reduces but does not eliminate this
• Capacitance: Parasitic capacitance between turns and to shield affects high-frequency behavior
• AC-DC difference: The relative difference between AC and DC resistance, (R_AC − R_DC)/R_DC, typically specified in ppm at stated frequencies

Calculable resistors with precisely known AC-DC difference serve as transfer standards between DC and AC resistance measurements.

Current from Voltage and Resistance

The most common approach to precision current measurement derives current from calibrated voltage and resistance:

• I = V/R: Measuring the voltage across a known resistance determines the current
• Series resistor method: Insert a precision resistor in the current path and measure the developed voltage
• Uncertainty combination: Current uncertainty includes contributions from both voltage and resistance uncertainties
• Self-heating: The measurement resistor must handle the measurement current without significant self-heating

This indirect approach transfers the accuracy of voltage and resistance standards to current measurements.

Current Comparators

Current comparators provide an alternative method for comparing currents with high precision:

• DC current comparators: Use magnetic cores and multiple windings to compare two currents, adjusting one until the net ampere-turns, and so the magnetic flux in the core, are zero; the current ratio then equals the inverse of the turns ratio
• Cryogenic current comparators: Superconducting shields and sensors achieve the highest sensitivity, enabling ratio measurements at parts in 10⁹ or better
• Flux detection: SQUIDs (superconducting quantum interference devices) or other sensitive magnetometers detect residual flux when currents are not perfectly balanced

Current comparators are essential tools for resistance scaling and for comparing standard resistors against the quantum Hall resistance.

Single-Electron Current Standards

Emerging technology aims to create current standards based on counting individual electrons:

• Single-electron pumps: Devices that transfer exactly one electron per cycle, producing a current I = ef, where e is the elementary charge and f is the pumping frequency
• Small currents: Even at a pumping frequency of 1 GHz, I = ef is only about 160 pA, far below the currents most calibrations need
• Quantum metrological triangle: Comparing a single-electron current against a Josephson voltage and a quantum Hall resistance closes the triangle of quantum electrical standards, a consistency test of the relations K_J = 2e/h, R_K = h/e², and I = ef
• Current status: Single-electron devices are approaching the accuracy needed for primary standards but have not yet achieved routine metrological use

Single-electron pumps would realize the ampere directly from its definition in terms of the elementary charge. A complementary route, already demonstrated in the laboratory, generates larger currents from the other two quantum standards: a programmable Josephson voltage drives a quantum Hall resistance, and a cryogenic current comparator amplifies the resulting current, I = V/R, to useful levels while preserving its traceability to K_J and R_K.

Practical Current References

Working current references for calibration purposes typically use:

• Voltage reference plus resistor: A stable voltage reference drives current through a precision resistor
• Current source ICs: Integrated circuits providing a stable output current set by an internal reference; traceability comes from calibrating that output
• Calibrated shunts: Precision resistors designed for current measurement, calibrated for their exact resistance value

Primary to Transfer Standard

Transfer standards receive their values directly from primary standards:

• Comparison method: Precision null measurements compare the transfer standard to the primary
• Multiple comparisons: Statistical analysis of multiple comparisons establishes the value and uncertainty
• Environmental recording: Temperature, pressure, and other conditions are recorded for correction
• Certificate issuance: Formal documentation of the assigned value and uncertainty

Transfer Standard Characteristics

Good transfer standards must be stable during transport and use:

• Transportability: Rugged enough to survive shipping without value changes
• Stability: Maintains value between calibrations; drift rate should be well characterized
• Temperature behavior: Known temperature coefficient enables correction for operating conditions
• History: Long calibration history demonstrates reliability and drift behavior

Zener voltage references and wire-wound resistance standards are common transfer standards for electrical measurements. A typical Zener transfer standard houses a temperature-controlled reference in a battery-powered enclosure, so that it stays powered in transit, and provides 10 V and 1.018 V outputs, the latter for continuity with the Weston cells such standards replaced.

Working Standards

Working standards are used for routine calibrations and are calibrated against transfer standards:

• Frequent use: Working standards bear the brunt of daily calibration work
• Multiple units: Laboratories maintain multiple working standards for redundancy and cross-checking
• Regular calibration: More frequent calibration than transfer standards to ensure continuing accuracy
• Known uncertainty: Working standard uncertainty includes drift allowance between calibrations

Calibration Hierarchy

The complete hierarchy from quantum standards to field instruments includes:

1. Intrinsic quantum standards: Josephson and quantum Hall effects (parts in 10¹⁰ or better)
2. National reference standards: Maintained at national metrology institutes (NMIs), calibrated against quantum standards (parts in 10⁸ to 10⁹)
3. Transfer standards: Portable references for interlaboratory comparisons (parts in 10⁷ to 10⁸)
4. Laboratory working standards: Used for customer calibrations (parts in 10⁶ to 10⁷)
5. Field instruments: Customer equipment calibrated against working standards (parts in 10³ to 10⁶)

Each step adds uncertainty, but the hierarchy enables practical distribution of traceability to millions of instruments worldwide.

International Comparisons

National metrology institutes participate in international comparisons to ensure worldwide consistency:

• Key comparisons: Organized under the CIPM Consultative Committee for Electricity and Magnetism, by the regional metrology organizations, and by the BIPM itself, these compare national standards to establish international equivalence
• Traveling standards: Stable transfer standards circulate among participating laboratories; for the quantum standards, the BIPM also carries its own Josephson and quantum Hall systems to national laboratories for on-site comparisons
• Published results: Comparison results are publicly available in the BIPM key comparison database (KCDB), documenting each NMI's degree of equivalence
• Mutual recognition: The CIPM Mutual Recognition Arrangement, signed in 1999, enables acceptance of calibration certificates worldwide

Null Measurement Methods

Precision comparisons typically use null methods where the detector indicates zero when the unknown equals the reference:

• Potentiometric method: Balance the unknown voltage against a known voltage derived from the standard
• Bridge methods: Resistance bridges balance the unknown against standard resistors
• Null detector sensitivity: The detector only needs to detect small imbalances, not measure absolute values
• Reduced detector errors: At null, detector gain and linearity errors do not affect the measurement

Thermal EMF Control

Temperature differences between junctions of dissimilar metals generate spurious voltages that can exceed the measurement resolution:

• Copper-copper connections: Use copper wires and terminals throughout low-level voltage circuits
• Thermal equilibrium: Allow time for connections to reach uniform temperature
• Draft protection: Shield connections from air currents that cause temperature gradients
• Reversal techniques: Reverse connections and average readings to cancel thermal EMFs
• Low-thermal connectors: Special connector designs minimize thermoelectric effects

Leakage and Guarding

Insulation resistance between conductors allows leakage currents that introduce errors:

• Guard shields: Driven shields intercept leakage current before it reaches sensitive nodes
• Triaxial connections: Inner conductor carries signal, middle conductor is guard, outer conductor is shield
• Clean surfaces: Contamination reduces surface insulation resistance; cleanliness is essential
• Low humidity: Moisture reduces insulation resistance; control humidity or use hermetic enclosures

Averaging and Statistics

Multiple measurements improve precision by reducing random noise effects:

• Allan variance: Statistical technique that characterizes stability versus averaging time
• Type A uncertainty: Standard deviation of the mean quantifies random uncertainty
• Outlier detection: Statistical tests identify anomalous readings for investigation
• Trending: Long-term records reveal systematic drift and environmental correlations

Calibration Laboratory Requirements

Laboratories performing traceable calibrations must meet specific requirements:

• Accreditation: ISO/IEC 17025 accreditation demonstrates technical competence
• Environmental control: Temperature, humidity, and contamination must be controlled within specified limits
• Equipment calibration: All reference standards and measurement equipment must have current calibration
• Uncertainty budgets: Documented uncertainty analysis for each measurement capability
• Quality system: Procedures, records, and management oversight ensure consistent performance

Industrial Calibration

Manufacturing and process industries require traceable calibration of working instruments:

• Working standards: Precision references calibrated against laboratory standards
• Calibration intervals: Defined periods between recalibrations based on stability and requirements
• Calibration records: Documentation of all calibrations for traceability and quality purposes
• Out-of-tolerance procedures: Actions taken when instruments are found outside specifications

Field Calibration

Calibration in field environments presents unique challenges:

• Portable standards: Rugged transfer standards that maintain accuracy during transport
• Environmental compensation: Correction for ambient temperature and other conditions
• Simplified procedures: Efficient methods suitable for field conditions
• Documentation: Complete records even under field conditions